Comparison of Pass Line Bets

This was a comparison I made between bets I've used, except for one or two. I used the WinCraps program. The PL is my starting betting point. That's my style. I've run some tests on these before & posted them on this board. However, a few problems became apparent, so I redid the whole thing. The betting styles are these, assuming a $5 table & $200 buy in: (1) PL only; (2) PL fullx; (3) PL, press c/o once, full; (4) PL 2x, 1 Come 2x; (5) Same, odds on; (6) PL full, 1C full; and, (7) PL full, Place 6&8 @ $12.

Results? At end of 240 rolls (a 2-hr. estimate), they came in:

1. PL full
2. PL Press c/o
3. PL only
4. PL full, 1C full
5. PL 2x, 1C 2x
6. Same, odds on
7. PL full, Place 6 & 8 @ $12

Results were different at the 1/2 hr. & 1 hr. times, times I'd think of walking.

PL full won an average of $45.20 over 10 2-hr. sessions, PL press won $27.00, "PL only" won $21.00, and PL full, 1 Come full won $9.50. All others lost, some big like (7): $64.70. One betting style is to press wins on the come out, so I included that. To bore you more, here's my methodology:

Each betting style was run for 10 2-hr. (240 roll) sessions with a different RNG number. I wanted to get a variety of roll sequences & having a different RNG number for each session ensured a different roll sequence. However, the RNG nos. did repeat for each betting style, so there was uniformity re: how a betting style compared w/ another on the same roll sequence. The RNG nos. were in 2 sets to simplify things, so there were 5 sessions w/ one set of nos., & 5 w/ a different set. But, all betting styles were run with the same RNG nos. for comparison purposes.

IOW, for example, I ran "PL only" with the RNG nos. 12321, 22321....52321; I ran the style a second time w/ RNG nos. 987654, 987655....987658. 2 sets of 5 nos. for the 10 sessions. Then, I ran the next style the same way & so on. So, "PL only," PL full, etc., each were run w/ the same set of RNG nos., if that makes any sense. I changed one digit in the RNGs to speed up things & these nos. were arbitrary. The graphing print-out allowed me to tally the 60, 120, 240 roll results.

Quote: Sonny44

PL full won an average of $45.20 over 10 2-hr. sessions,

not really sure why you only ran 10 sessions for each betting system.
you could have run 10 2-hr. sessions 10,000 times to get a better view of the averages one could easily come up against and the range around those averages.

here is data from 10 million 240 roll sessions
showing just the pass/dpass with some odds sprinkled in as I do not see what full odds you actually used.
The average 240 roll session has a net loss

any few sessions could produce a winning multiple session, actually just pausing the one lifetime session.

40uBR-240rolls	Avg. No. games	Avg. No. dice rolls	Bankroll busted	Bankroll decreased	Bankroll increased	Avg end bankroll net	SD ending bankroll
pass 0 odds	71.84	242.52	23	52.74%	42.58%	-1.02	8.46
dpass 0 odds	69.85	242.52	68	51.92%	43.32%	-0.98	8.37
pass 1x odds	71.67	241.93	1.30%	52.70%	47.05%	-1.01	16.01
dpass 1x odds	69.62	241.72	1.59%	52.02%	47.74%	-0.98	15.96
pass 2x odds	69.74	235.42	9.90%	51.90%	47.77%	-0.98	23.79
dpass 2x odds	67.62	234.81	10.16%	51.17%	48.51%	-0.95	23.68
pass 345x odds	61.8	208.63	32.15%	53.63%	45.57%	-0.86	37.91
dpass 345x odds	61.01	211.84	29.80%	53.34%	45.87%	-0.87	37.54

Imagine being one of the line bettors flat betting busting out of a 40 unit bankroll in 2 hours of play!
it could happen

btw, your example of a $200 bankroll and $5 bets = 40 betting units (200/5)
the same ratio as one buying in for $4000 and betting with $100 chips or one buying in for $400 and betting $10 units.

For some, it is fun to see what could happen in any session using simulations.
of course, many do not see a rng dice roll being equal to a real dice roll on a real craps table,
one could also run simulations against a large sample of actual dice rolls too.
one could also just do the math to see how accurate the simulation results are and vice versa. (probabilities, averages and variances)

that could be fun two
sometimes it is more fun with three

or in threes

of course
Two Out of Three Ain't Bad
(just another love song)

Sally

Quote: mustangsally

not really sure why you only ran 10 sessions for each betting system.

for sure, never lose sight of the really large number of rolls you can do with Wincraps ... and should do if you are trying to come to conclusions.

Quote: Sonny44

Each betting style was run for 10 2-hr. (240 roll) sessions

when you do start doing more, I think you will realize right away this is insufficient. Wincraps will cure you.

I guess I mistakenly ran these tests according to how much time I plan to stay at a table, to make the sim more realistic. Perhaps I will go back & do 10K rolls. Maybe that's more realistic in comparing these bets.

Quote: Sonny44

I guess I mistakenly ran these tests according to how much time I plan to stay at a table, to make the sim more realistic. Perhaps I will go back & do 10K rolls. Maybe that's more realistic in comparing these bets.

Sonny,
I run my sessions the same way. We already know what the outcome will be if you run 100,000 rolls. What I like to know is what could happen in a 2-4 hour session, given my buyin amount, and my betting style for that day. I run 100 sessions and then view the outcomes of those sessions. If the results are really skewed, I will run 100 games a few more times.

I think understanding how to use standard deviation would also give you some idea of the extent of possible outcomes, but our way also shows some possible results. Just make sure you understand that even though you may see great results in WinCraps on this small sample, it doesn't mean you will ever see the result for real.

Quote: RaleighCraps

Sonny,Just make sure you understand that even though you may see great results in WinCraps on this small sample, it doesn't mean you will ever see the result for real.

Thanks, RC. Good to know someone else is doing the same thing. Yes, what you say here is very true. I played on a $10 table w/ a $300 buy in. I went PL 3/4/5 from the start, which was the best betting style in my sims. In 20 min., I was broke. I still think it's the best of the styles I simmed, but I have to use my head, such as start small then go from there, depending on health of my br.

Quote: Sonny44

I guess I mistakenly ran these tests according to how much time I plan to stay at a table, to make the sim more realistic. Perhaps I will go back & do 10K rolls. Maybe that's more realistic in comparing these bets.

NO. a 240 roll session is fine for a sim run.
so is your other roll amounts

But only 10 sessions run for each?

Here is an example why you want to run 10,000 sessions of 240 rolls each
to get a better idea of the average and the variance and the results will be much closer to actual probabilities than just running 10 sessions.


Sally 10 coin flip example
I simulated flipping a coin 10 times... yep,
and I ran 10 sessions

here is the histogram
It PROVES that 6 heads is more common than any other outcome.
I am the GREATEST!!!
case is now closed
Sally IS the smartest human EVER!!!


below shows what the histogram looks like after 1 million 10 coin flip sessions
I had a feeling that just maybe, because I could never flip that many coins (10 million coin flips)
maybe
maybe


I like the first pic the best
The second one looks like it came out of a text book and has nothing to do with reality

But some may say it does
because my sample size was way larger than the first sim (1 million sessions verses only 10 sessions)
Hmmm

Sally

Quote: RaleighCraps

I run 100 sessions and then view the outcomes of those sessions. If the results are really skewed, I will run 100 games a few more times.

Just run 10,000 sessions each, be done and you can have more faith in the results than just 10 or 100 sessions
This should be a given
Sally

Quote: Sonny44

I played on a $10 table w/ a $300 buy in. I went PL 3/4/5 from the start,

300/10 = 30 bankroll units
My sim data was for 40 units and one can see how high the bust rate is taking 345x odds.

One can also see how much is lost on average per session played.

Small bankrolls and High odds = high bust rates
be prepared... no way to get around it.

Sally

Quote: mustangsally

300/10 = 30 bankroll units
My sim data was for 40 units and one can see how high the bust rate is taking 345x odds.

One can also see how much is lost on average per session played.

Small bankrolls and High odds = high bust rates
be prepared... no way to get around it.

Sally

Yes, all you say is true & I learned my lesson. I need to be able to read a table more & bet accordingly. I think the main measure is health of my rack. But, I can't stay at a table for 10k+ rolls. All I was doing was comparing these betting styles in real life situations. That's what was most important to me, rather than getting some kind of ultimate statistical read on all the styles.

It could be true, tho, that one of the other styles, over 10k+ rolls, might prove better. Also, experienced craps players like Frank S. & the Wizard have said PL w/ odds, is the best craps bet. The only thing better, Wiz said, is the DP w/ odds. Of course, that doesn't guarantee the dice will roll in my favor.

Quote: mustangsally

Just run 10,000 sessions each, be done and you can have more faith in the results than just 10 or 100 sessions
This should be a given
Sally

No need to run 10,000 sessions though. It will show what we already know, and that is the vast majority of the sessions will all be losers.
If I run 100 sessions with a really bad system, and 99 of them lose, well that tells me I have no chance with that system.
If I run 100 sessions, and 60 of them are winners, well I have an idea that I have a 50-50 chance of winning. Of course, it could be the RNG just happened to give me the 100 session rolls that made the system look good, when in fact, it would lose 95% of the time in 10,000 sessions. But once again, they all lose in 10,000 sessions.

The one place I have found where running large sessions really helps, is if you want to see the affect of different size bankrolls for the same system. In other words, play PL w/10x odds and playing $135 across, and do it with a $1K BR, a $2K BR, and a $5K bankroll. The $1K BR will bust A LOT. The $2K busts out less often, and you end up with a few more winning sessions. The $5K busts out the least, and in my sequence of rolls, had the most winning sessions, but you end up losing 5x as much as the $1K BR each time it busts. But it did show me how important BR size is compared to what your bet(s) are going to be.

In the end though, the only thing that really matters is what the dice are doing at the time you are playing. You can be playing the dumbest bets on the table (the HORN), but if 2,3,11, and 12 are coming up a lot, you are making money, and everyone else will be losing.

My data showed me that buying in for $700 and playing $135 across is pure BR suicide, yet that is exactly what I did last weekend. Four straight sessions I won money, two of them less than $200, and the other two were over $1800 each. WinCraps would show that as almost impossible, yet it happened. Gotta love craps.

I run short sessions on Wincraps all the time for the hell of it. But if I'm trying to come to conclusions, gotta run big numbers.

Quote: odiousgambit

I run short sessions on Wincraps all the time for the hell of it. But if I'm trying to come to conclusions, gotta run big numbers.

Well, big numbers are fine. To me, I only play in sessions for an hr. or two, if lucky. But, it doesn't matter if I'd play 8, 10 hrs. What this means is that the particular sequence of dice during that session is only a small clip out of the 10k, etc. roll tape of the dice. And, there are probably an infinity of small clips out of that grand sequence.

And, if the RNG seed is changed for each 10k roll, then a different grand sequence is generated, resulting in different results. What I'm saying is however many changes of an RNG seed (which could be infinite), there are an infinite number of different "sessions" within all those grand sequence tapes.

Maybe I'm off base, but this is how I'm seeing it. All I wanted to do, as I've said, is to compare these bets against each other, using the same dice sequence, using 2 different RNG seeds. Maybe it's not adequate in the grand scheme of things, but according to what I've said, I don't think any number of rolls is adequate. IOW, probabilities are fine, but there's little chance they will hold true for a session at a dice table.

The game is truly random, and all I wanted is some kind of indication of which of the bets I have in mind will produce the best results in the limited time frame of a session. In the end, it boils down to whether I'm lucky or not, as I've described in my experience at a $10 table w/ a $300 buy in.

Quote: Sonny44

All I wanted to do, as I've said, is to compare these bets against each other, using the same dice sequence, using 2 different RNG seeds. Maybe it's not adequate in the grand scheme of things, but according to what I've said, I don't think any number of rolls is adequate.

1 million rolls will be way more adequate than 240 rolls. But that is NOT the issue here.

You want a sim of 240 dice rolls but only want to run 10 sessions (or 10 different players playing)
instead of 10 million of them (as I ran).
and you say your results and mine are both meaningless as anything can happen in any session at Craps.
this shows your lack of understanding the law of Large Numbers.
it is your friend.

It is the same as saying 10 players played your system and all won, you are a GOD
will 10 million players all win by doing the same?
NO
way

By running ONLY 10 sims of 240 rolls (your session length) the data from only those 10 sessions is really absolutely meaningless.

Just like my coin flip example where I PROVED that in 10 coin flips, 6 heads is more likely than any other outcome.

The larger the SAMPLE SIZE, the more accurate the results will be (as long as the code is doing as programmed)

Quote: Sonny44

The game is truly random, and all I wanted is some kind of indication of which of the bets I have in mind will produce the best results in the limited time frame of a session.

just doing the math will tell you that. expected value and variance are friends here too

the problem is your 10 sessions only produced results that if done another 10 sessions could easily be way different from each other.
Your 10 session sample size is WAY TOO SMALL.
Size matters in running simulations.

Quote: Sonny44

In the end, it boils down to whether I'm lucky or not, as I've described in my experience at a $10 table w/ a $300 buy in.

NO.
It boils down to every possible outcome weighted by the probabilities of that outcome happening.
average and variance are friends too.

In other words, each session boils down to the math of the game.
Know what to expect.
if you think you can win more times than expectation on every bet you make, keep that a secret as you will beat the game until you die.

thoughts...
It is not luck that the most common roll for a 7out to happen on by any shooter is roll #3
It is NOT LUCK that the 2nd most common roll for a shooter to 7 out on is roll #2.

It is the math of the game.
The math is a friend too


at this point, all your gathered data, in my opinion, is still meaningless and a waste of effort (for data not knowledge in running a sim)
that should be followed up by larger sim runs.

240 rolls is fine for ONE session length
now run 10,000 sessions
each and post your results

(pretend that 10,000 other players are trying out your systems and only you (not any of them)
want to see how many win their session, lose and what the average over all of them is and the range around that average.)
WinCraps tracks all the data for us.

Sally

Quote: RaleighCraps

No need to run 10,000 sessions though. It will show what we already know,

just like my coin flip example
some will never "get it" and that is fine too
I did that experiment (simulation) again, but this time 100 coin flips and 10 sessions (1000 total coin flips)
shows the same result as did my 10 flip experiment
50 heads is NOT the most common number of heads in 100 flips
51 IS
see


size matters in running simulations

100 sessions of 2 to 4 hours gets that ball rolling but
10,000 sessions shows a more accurate and complete picture. (I am guessing that your computer is slower than mine.)

Quote: RaleighCraps

and that is the vast majority of the sessions will all be losers.

not even true
show your proof of this please

Quote: RaleighCraps

In the end though, the only thing that really matters is what the dice are doing at the time you are playing. You can be playing the dumbest bets on the table (the HORN), but if 2,3,11, and 12 are coming up a lot, you are making money, and everyone else will be losing.

this may be true for one session played. but the math or sim data will say
"Hey, you only want a 24% chance of winning and
you want to win 15 sessions in a row or just 13 out of 15?!

GOOD LUCK!

Quote: RaleighCraps

My data showed me that buying in for $700 and playing $135 across is pure BR suicide, yet that is exactly what I did last weekend. Four straight sessions I won money, two of them less than $200, and the other two were over $1800 each. WinCraps would show that as almost impossible, yet it happened. Gotta love craps.

Not even close to the truth, but cool you shared your opinion here.

show your proof of this if you think that simulation size does not matter

It may show you have a 30% chance of winning over one session length played.

30%*30%*30%*30% = ?
NOT almost impossible

any player could do it, but not almost all of them will and
not consistently. like 2 out of 3 or 3 out of 4
4 group sessions

and your 4 sessions were only really one session played, the lifetime session that only gets paused each time you stop playing
I mean really, place bets
are the main sucker bets at craps over any lifetime session.
about 1 in a million, on average, Craps players over any lifetime session will show a net profit from making so many place bets.
not great odds here
are YOU that one because of money management and bet selection (when to increase, decrease and remove those nasty place bets)

That is cool too.
1 out of a million... More money for me to win from the casinos!
yes!!

place bets, oh
the wait for the monster roll to have a big win to cover all the smaller losses is what the craps place bettor says and believes 100%
that is cool too. more fun hoping that way as I see it

the Law of Large Numbers comes into better focus the longer one keeps playing craps.
and one can not get around it.
it is our friend

have more fun making all those place bets

Sally

gotta defend R.Craps, seems to me you are pouncing on words like "suicide" here Sally, whereas he is just using hyperbole. I don't think he would disagree with what you go on to say.

The OP, I dunno. I was there once, at least I can say that I felt 10,000 dice rolls was sufficient at one time. I know better now.

Sally I have a few question for you, the first one is very simple, how often do you play craps in a real casino?

The second one, please tell me how the Law of Large Numbers is our friend, after all we are not the casino!

Quote: MustangSally

thoughts...
It is not luck that the most common roll for a 7out to happen on by any shooter is roll #3
It is NOT LUCK that the 2nd most common roll for a shooter to 7 out on is roll #2.

How do the two statements above fit in with the average number of rolls for a player being 8.5 according to The Wizzard?

https://wizardofodds.com/games/craps/number-of-rolls/

I am sure I am missing a connection some place that would let me understand how they are related, but at this point, I cannot figure it out.

Quote: odiousgambit

gotta defend R.Craps, seems to me you are pouncing on words like "suicide" here Sally, whereas he is just using hyperbole. I don't think he would disagree with what you go on to say.

The OP, I dunno. I was there once, at least I can say that I felt 10,000 dice rolls was sufficient at one time. I know better now.

Thanks OG for the defense, but Sally is probably more right here than I am. I am assuming she can provide proof of what she writes, and I am quite confident that her math skills would allow her to do so if asked, but I am not going to ask.

I just believe that the game of craps, when being played for fun and perhaps with some reckless abandon, can in fact, be beaten some of the time, with some bets that are less favorable than what the math says must happen over the Law of Large Numbers. That very statement is blasphemy to any self respecting mathematician, and so I do not expect to ever change their view. And there is NO WAY I can prove that I might be right, and the math just doesn't work for my lifetime of craps.

I appreciate the fact that the math people have not given up on trying to get me to see the error of my ways, and I also appreciate the craps players who have played with me, have seen me ahead more than 2x my buy-in on a number of occasions, and come to my defense on some of my thoughts.

My blogs and craps stories are real. My losses are real, AND reported. My wins are reported. I have nothing to gain by writing fiction. I may very well just be running over 1 SD to the good. All I know is, I have had a large number of big wins, making the bad Place and Buy bets, and I have done it when people were playing the 'BEST' way (PL w/ full odds) and they got crushed. It has happened with some of our own forum members playing along side of me. Very infinitesimal in the law of Large Numbers, but very real in my wallet that day and theirs. And since I play in the hopes to win big money, I'll take my wins now. After all, who knows, I may just stop playing, which means the craps game will never get enough rolls from me to get even with me for playing so stupidly.

Quote: RaleighCraps

Quote: MustangSally

thoughts...
It is not luck that the most common roll for a 7out to happen on by any shooter is roll #3
It is NOT LUCK that the 2nd most common roll for a shooter to 7 out on is roll #2.

How do the two statements above fit in with the average number of rolls for a player being 8.5 according to The Wizzard?

https://wizardofodds.com/games/craps/number-of-rolls/

The 8.5255 comes from the sum of the probabilities of a shooter making Roll X
All shooters make roll1 and roll2
It goes down hill from there.
only 0.49721087
make roll7 - that is less than 50%

here is my table for this
it is correct that roll2 and roll3 are really close
They do fight it out for top spot, top dog, as the number of completed shooters increase.

Prob to make roll X	roll	7out on
1	1	0
1	2	0.111111111
0.888888889	3	0.116769547
0.772119342	4	0.104766804
0.667352538	5	0.091223629
0.576128909	6	0.078918038
0.49721087	7	0.068166764
0.429044107	8	0.058852758
0.370191349	9	0.05080065
0.319390699	10	0.043844137
0.275546562	11	0.037836136
0.237710426	12	0.032648501
0.205061925	13	0.028170022
0.176891903	14	0.024304335
0.152587569	15	0.020968008
0.13161956	16	0.018088857
0.113530703	17	0.015604454
0.097926249	18	0.013460839
0.08446541	19	0.011611381
0.072854029	20	0.0100158
0.062838229	21	0.008639309
0.05419892	22	0.007451869
0.046747051	23	0.006427548
0.040319503	24	0.005543963
0.03477554	25	0.004781795
0.029993744	26	0.004124373
0.025869371	27	0.00355731
0.022312061	28	0.003068195
0.019243866	29	0.002646317
0.01659755	30	0.002282437
0.014315113	31	0.001968585
0.012346528	32	0.001697884
0.010648644	33	0.001464404
0.009184241	34	0.001263027
0.007921214	35	0.00108934
0.006831874	36	0.000939536
0.005892338	37	0.000810332
0.005082006	38	0.000698895
0.004383111	39	0.000602782
0.003780328	40	0.000519887
0.003260442	41	0.000448391
0.002812051	42	0.000386727
0.002425324	43	0.000333543
0.002091782	44	0.000287673
0.001804109	45	0.000248111
0.001555998	46	0.000213989
0.001342009	47	0.000184561
0.001157448	48	0.000159179
0.00099827	49	0.000137288
0.000860982	50	0.000118407
0.000742575	51	0.000102123
0.000640451	52	8.80787E-05
0.000552373	53	7.59656E-05
0.000476407	54	6.55184E-05
0.000410889	55	5.65079E-05
0.000354381	56	4.87366E-05
0.000305644	57	4.20341E-05
0.00026361	58	3.62533E-05
0.000227357	59	3.12675E-05
0.000196089	60	2.69674E-05
0.000169122	61	2.32587E-05
0.000145863	62	2.006E-05
0.000125803	63	1.73012E-05
0.000108502	64	1.49219E-05
9.35801E-05	65	1.28697E-05
8.07104E-05	66	1.10998E-05
6.96106E-05	67	9.57328E-06
6.00373E-05	68	8.25671E-06
5.17806E-05	69	7.12119E-06
4.46594E-05	70	6.14184E-06
3.85176E-05	71	5.29718E-06
3.32204E-05	72	4.56868E-06
2.86517E-05	73	3.94036E-06
2.47114E-05	74	3.39846E-06
2.13129E-05	75	2.93108E-06
1.83818E-05	76	2.52798E-06
1.58538E-05	77	2.18032E-06
1.36735E-05	78	1.88047E-06
1.17931E-05	79	1.62185E-06
1.01712E-05	80	1.39881E-06
8.77239E-06	81	1.20643E-06
7.56596E-06	82	1.04052E-06
6.52544E-06	83	8.97419E-07
5.62802E-06	84	7.74E-07
4.85402E-06	85	6.67555E-07
4.18646E-06	86	5.75749E-07
3.61072E-06	87	4.96568E-07
3.11415E-06	88	4.28277E-07
2.68587E-06	89	3.69378E-07
2.31649E-06	90	3.18579E-07
1.99791E-06	91	2.74766E-07
1.72315E-06	92	2.36978E-07
1.48617E-06	93	2.04387E-07
1.28178E-06	94	1.76279E-07
1.1055E-06	95	1.52036E-07
9.53468E-07	96	1.31127E-07
8.22341E-07	97	1.13094E-07
7.09248E-07	98	9.75402E-08
6.11708E-07	99	8.41259E-08
5.27582E-07	100	7.25563E-08
4.55025E-07	101	6.25779E-08
3.92447E-07	102	5.39718E-08
3.38476E-07	103	4.65493E-08
2.91926E-07	104	4.01475E-08
2.51779E-07	105	3.46262E-08
2.17153E-07	106	2.98642E-08
1.87288E-07	107	2.57571E-08
1.61531E-07	108	2.22148E-08
1.39317E-07	109	1.91597E-08
1.20157E-07	110	1.65247E-08
1.03632E-07	111	1.42521E-08
8.938E-08	112	1.22921E-08
7.70879E-08	113	1.06016E-08
6.64863E-08	114	9.14361E-09
5.73427E-08	115	7.88613E-09
4.94566E-08	116	6.80158E-09
4.2655E-08	117	5.86618E-09
3.67888E-08	118	5.05943E-09
3.17294E-08	119	4.36362E-09
2.73658E-08	120	3.76351E-09
2.36023E-08	121	3.24593E-09
2.03563E-08	122	2.79953E-09
1.75568E-08	123	2.41452E-09
1.51423E-08	124	2.08246E-09
1.30598E-08	125	1.79607E-09
1.12637E-08	126	1.54906E-09
9.71469E-09	127	1.33602E-09
8.37866E-09	128	1.15229E-09
7.22638E-09	129	9.93816E-10
6.23256E-09	130	8.5714E-10
5.37542E-09	131	7.39261E-10
4.63616E-09	132	6.37593E-10
3.99856E-09	133	5.49908E-10
3.44866E-09	134	4.74281E-10
2.97438E-09	135	4.09055E-10
2.56532E-09	136	3.52799E-10
2.21252E-09	137	3.0428E-10
1.90824E-09	138	2.62433E-10
1.64581E-09	139	2.26342E-10
1.41947E-09	140	1.95214E-10
1.22425E-09	141	1.68367E-10
1.05589E-09	142	1.45212E-10
9.10674E-10	143	1.25242E-10
7.85432E-10	144	1.08018E-10
6.77415E-10	145	9.31623E-11
5.84253E-10	146	8.03501E-11
5.03902E-10	147	6.92998E-11
4.34603E-10	148	5.97693E-11
3.74833E-10	149	5.15494E-11
3.23284E-10	150	4.446E-11
2.78824E-10	151	3.83457E-11
2.40478E-10	152	3.30721E-11
2.07406E-10	153	2.85237E-11
1.78882E-10	154	2.46011E-11
1.54281E-10	155	2.12177E-11
1.33064E-10	156	1.82998E-11
1.14764E-10	157	1.5783E-11
9.89809E-11	158	1.36124E-11
8.53684E-11	159	1.17404E-11
7.3628E-11	160	1.01258E-11
6.35022E-11	161	8.73324E-12
5.4769E-11	162	7.5322E-12
4.72368E-11	163	6.49625E-12
4.07405E-11	164	5.60285E-12
3.51376E-11	165	4.83236E-12
3.03053E-11	166	4.16778E-12
2.61375E-11	167	3.59457E-12
2.25429E-11	168	3.1003E-12
1.94427E-11	169	2.67386E-12
1.67688E-11	170	2.30616E-12
1.44626E-11	171	1.98896E-12
1.24737E-11	172	1.71552E-12
1.07582E-11	173	1.47948E-12
9.27866E-12	174	1.27609E-12
8.0026E-12	175	1.10056E-12
6.90204E-12	176	9.49241E-13
5.95282E-12	177	8.18678E-13
5.13415E-12	178	7.05991E-13
4.42807E-12	179	6.09068E-13
3.8191E-12	180	5.25135E-13
3.29387E-12	181	4.53082E-13
2.84088E-12	182	3.90687E-13
2.45018E-12	183	3.36953E-13
2.11322E-12	184	2.90656E-13
1.82259E-12	185	2.50577E-13
1.57194E-12	186	2.1616E-13
1.35576E-12	187	1.86517E-13
1.16931E-12	188	1.6076E-13
1.00849E-12	189	1.38778E-13
8.698E-13	190	1.19571E-13
7.5018E-13	191	1.0314E-13
6.47011E-13	192	8.90399E-14
5.5803E-13	193	7.67164E-14
4.81286E-13	194	6.61693E-14
4.15096E-13	195	5.70655E-14
3.5801E-13	196	4.92939E-14
3.08774E-13	197	4.24105E-14
2.6631E-13	198	3.66374E-14
2.29685E-13	199	3.16414E-14
1.98097E-13	200	2.72005E-14
1.70854E-13	201	.
8.525510204	total	.

Sally

Quote: mustangsally

Quote: RaleighCraps

My data showed me that buying in for $700 and playing $135 across is pure BR suicide, yet that is exactly what I did last weekend. Four straight sessions I won money, two of them less than $200, and the other two were over $1800 each. WinCraps would show that as almost impossible, yet it happened. Gotta love craps.

Not even close to the truth, but cool you shared your opinion here.

show your proof of this if you think that simulation size does not matter

It may show you have a 30% chance of winning over one session length played.

30%*30%*30%*30% = ?
NOT almost impossible

Is the 30% pulled out of the air, or were you able to calculate this as a real number? I am not trying to be a jerk, I am really interested in the answer. The reason I am asking is, from the Sims I have run over the years, my gut is telling me if I am betting $80 across, $10 PL, and $100 odds, on a $700 buy-in, I am betting such a large percentage of my buy-in, that I should be doomed to failure much more often than 70% of the time. And when I increase that to $135 across because I am seeing 4s and 10s being rolled, that puts even more pressure on that $700 buy in.

At your 30%, I would come out ahead 4 sessions in a row (.3x.3x.3x.3) = .81% of the time or less than 1 out of the next 100 sessions I play. That is 3 or 4 years to complete. Looks pretty damn small to me.

However, I think 30% is too high, betting over 1/3 of my buy-in. My gut says the failure rate should be more around 85-90% of the time. Let's use 85%. (.15X.15x.15x.15)= .00506% of the time. I might screw this up, but I think that says it would happen 1 in 2000 sessions? That is going to take me 15 years to complete, and I'd be shocked if I get 15 years more of craps play, for any number of reasons.

Quote: mustangsally

any player could do it, but not almost all of them will and
not consistently. like 2 out of 3 or 3 out of 4
4 group sessions

and your 4 sessions were only really one session played, the lifetime session that only gets paused each time you stop playing
I mean really, place bets
are the main sucker bets at craps over any lifetime session.
about 1 in a million, on average, Craps players over any lifetime session will show a net profit from making so many place bets.
not great odds here
are YOU that one because of money management and bet selection (when to increase, decrease and remove those nasty place bets)

That is cool too.
1 out of a million... More money for me to win from the casinos!
yes!!

Because the bets are not static, but are dynamically being adjusted, sometimes correctly, and sometimes not, I maintain that my 4 sessions are important. Yes, I understand your point that starting and stopping has no effect on my ONE Lifetime session, BUT if I have an artificial stopping point that says, "when I am ahead, I walk", that has to have an effect on the lifetime session.

After all, when I restart I will not be betting the same total amount I was betting when I quit. I will be restarting at a lower amount. When the game is rolling good, the amounts bet are larger, and when it is working, I have my biggest bets out there when I seem to be winning more than I am losing. Totally unscientific, but yet, it happens. That is how I win 4x my buy-in. It is also how I lose my total buy-in when things just don't go my way.

I am certainly not trying to say Play Across is the way to riches. However, I do not agree that playing PL w/full odds, and then making CB w/full odds is 'the best' way to play, which is often stated here. The math may say that is so, but the real life craps game does not follow the math for that one session. (Yes, I know we are now back to the, Lifetime is one session, Large Number Law, etc.). And many who accept this way of play are grossly underfunded on their buy-in amount to survive playing this way.

I tried playing strictly PL/odds and CB/odds, and I have gotten crushed too often. Come out 7s kept wiping out my CBs, and if I had the odds working, would virtually guarantee me I would lose on this shooter. Playing 1 PL bet with full odds is boring as heck, and requires the shooter to make a point. Hardways are sort of fun, but overall, they are pretty boring too most of the time. Horn is just stupid. Hop bets are even worse.

So, I choose to Buy bet the 4/10 and Place bet the 6/8, both of which are not that far off in HE from a PL bet.
I agree that Place betting the 5/9 is pretty dumb, but how the dice know I am not on those numbers I have no idea. But if I leave those numbers open, I guarantee you I will see so many 5s and 9s you will call for a dice inspector (hyperbolic statement). Fortunately, MS gives us a Buy on win for the 5/9, which makes the bet just slightly worse than the Buy 4/10.

Thanks once again for the engaging conversations. I am afraid we will never see eye to eye on how to play craps, but I hope to be able to play along side of you in a craps game a few times. It would be fun to see who makes out better on that small sample. Sure it is meaningless in the big picture, but when we leave for the day, one of us will have more money than the other, and that is not meaningless.

It is ironic that we have these conversations, as I drive a Mustang, and my wife's favorite song is Mustang Sally. Feels like a cosmic connection (is that a scientific reality?)
RC

Quote: mustangsally

Quote: RaleighCraps

Quote: MustangSally

thoughts...
It is not luck that the most common roll for a 7out to happen on by any shooter is roll #3
It is NOT LUCK that the 2nd most common roll for a shooter to 7 out on is roll #2.

How do the two statements above fit in with the average number of rolls for a player being 8.5 according to The Wizzard?

https://wizardofodds.com/games/craps/number-of-rolls/

The 8.5255 comes from the sum of the probabilities of a shooter making Roll X
All shooters make roll1 and roll2
It goes down hill from there.
only 0.49721087
make roll7 - that is less than 50%

Sally

Thanks. I knew there had to be a connection that I just wasn't understanding.

Quote: RaleighCraps

And since I play in the hopes to win big money, I'll take my wins now. After all, who knows, I may just stop playing, which means the craps game will never get enough rolls from me to get even with me for playing so stupidly.

I believe it was the famous economist, John Maynard Keynes, who said, referring to the stock market, "In the long term, we're all dead."

Quote: RaleighCraps

I am certainly not trying to say Play Across is the way to riches. However, I do not agree that playing PL w/full odds, and then making CB w/full odds is 'the best' way to play, which is often stated here.RC

Hope I'm not one of those indicated in the underlined words. In my testing (which may or not be adequate), PL full, 1 Come full; PL 2x, 1 Come 2x; Same w/ odds on, all significantly underperformed PL full odds.

I'm not smart enough to bet like you nor to write a script to run your betting method. I wish I were. But, it would be very interesting to test your method, using my testing methodology, to compare w/ the bets I ran. Even then, as you say, it's an up & down proposition, regardless of the math, given that our sessions, perhaps even our "lifetime sessions," are mere clips in that infinitesimal tape of dice rolls, which the mathematicians seem most concerned with.

Mustang Sally: For each bet, I ran 10 sessions of 240 rolls each. That means each bet was tested for 2,400 rolls. Not 10k, but certainly more than a mere 240. I wanted real-life parameters. Like I said, I just wanted some indication of which bet would perform best, not that the "winner" would guarantee wins at the table.

Sorry for the ambiguity Sonny. When I wrote the word here, I meant the forum, not your thread, or you.

This has been an ongoing discussion here for a couple of years now. The math folks will say run any bet you want for 1M rolls, and see what your loss is. The bet that loses the least amount is the best bet to make.

I say I am never going to bet 1M rolls. So, if I find a bet that pays me in spurts, it 'could' be better for me in my life's play of craps to play the higher risk bet, rather than then bet with the least amount of loss after 1M rolls.

I don't play to lose the least amount. I play to try and win large amounts.

In the end, the more we play, the higher the likelihood that we will lose money.

Here's the thing about the game. It all hinges on whether you happen to have the bets in the right place for the next throw of the dice. In the end, the game is really that simple.
If you are a $10 PL w/full odds bettor, but you happen to make a $500 hard 4 bet because Miss Cleo told you it was coming, and it hits? You are going to have a winning session that day, if you go back to how you normally bet. And it won't matter if people are making points, or not making points. You will be a winner because you had a large amount of money on one bet that really paid off.

I almost always play PL w/ full odds, if some points are being made. If I am on a table where no one is making points, that is when I will sometimes abandon PL betting, and just Place/Buy bet, or heaven forbid, go to the dark side.

The game is partly art. You have to be willing to change what you like to do, to adapt to how the table you are at seems to be running. A classic example was a session I had with another forum member at Green Valley Ranch. We were noticing a much larger number of 12s were being rolled, and we kept commenting on it. Finally, he got 30 white chips, and he bet the 12 for 30 straight rolls. It hit 4 times in those 30 rolls! Dumbest bet on the table, and he made money on it. 10 minutes later, and we did not see a 12 for close to 30 minutes.

RaleighCraps the trouble with the math of the game is it's used to see what the odds are for the casinos to see what they can rip you off for - legally, by not paying the correct odds, and to make sure that they will make money with every bet on the table! Now the casinos have to make money and they do it by not paying you the right amount when you win a bet, so we are always playing a negative game when we are playing craps.

Just like a computer you are always going to get bad data out of it if you are inputting bad data, i.e. GIGO, Garbage In Garbage Out.
Any time you are running a craps simulator like WinCraps you are always inputting bad data, because you are always leaving out the H equation of the formula you are running!

There is no long run while you are at the table, that what the math guys can't get through their heads, you will never be there for 10,000 rolls of the dice in one session, so therefore there is no long run. You are only there until your money runs out or you decide to leave.

Just because the math of the game says the pass-line bet is a good bet because of the low house edge, it doesn't mean that you will win those bets. As you know anything can happen in the short run, and when you are playing craps you are always playing in the short run. Your game ends when you leave the table, there is no other way of looking at it, you could be killed walking out of the casino when someone is in such a big hurry to get into the casino knocks you down by accident, and you hit your head. While you had every intention of coming back the next day and playing again you never reached the 10,000 roll mark. While the casinos game never ends, it goes on 24 hours a day seven days a week.
Their game is based on knowing that on every winning bet they are going to take out a Vig.

When you are playing in the short run and you are changing your bets as need be, there would have to be way more things to consider other then just the pass-line bet.
I see an awful lot of players losing when they are betting only on what are considered the three best bets to make, and that is betting on the 6 and 8 and taking full odds on whatever point has been established.

In the short-term those bets could be the worst bets on the table to make, because no one is rolling them!

Quote:

RaleighCraps

The game is partly art. You have to be willing to change what you like to do, to adapt to how the table you are at seems to be running. A classic example was a session I had with another forum member at Green Valley Ranch. We were noticing a much larger number of 12s were being rolled, and we kept commenting on it. Finally, he got 30 white chips, and he bet the 12 for 30 straight rolls. It hit 4 times in those 30 rolls! Dumbest bet on the table, and he made money on it. 10 minutes later, and we did not see a 12 for close to 30 minutes.

I think that I showed you personally, on two other occasions, that by making what some of the math guys consider bad bets, that you could make money.

Like when the so-called random roller showed up at GVR and I told you that has soon as the guy made one field number to put your money in the field, and parlay each hit you got for 4 or 5 rolls of the dice!
At the time you were losing, and if it wasn't for the fact that you had to catch a plane you would have came out a winner! As the guy was rolling 4 to 6 field numbers in a row, every time he made his first one. You asked how I knew to bet on that guy, because he was nothing but a so-called random roller, my answer was very simple, I've seen him on a bunch of different tables and that was what he did every time he had the dice! If you open your eyes when you are playing and you see different regular players doing the same thing over and over again, you can pick-up on a betting strategy that will work for that one shooter! Will it work every time, hell no, you have to use your head when betting on anybody.

I also showed you that you should always have the 4's and 10's bet when I'm shooting, and it very funny to watch the don't players always laying those numbers when I'm just killing them! Some players never learn if the shooter beats you once, you should lay off them!

The math of the game says that all of those bets are bad bets, but the problem is the dice do not know that, they don't know that they have to follow the math of the game, and that is why we can win when playing craps.

Without being able to put the H equation into WinCraps you're just putting useless data in,... GIGO.
I track the table when I'm playing but never use that data when I get home, all the math guys will tell you that what happened in the past has no effect on what is going to happen in the future.
So if that is true, then even if the math says that you have a good bet with the 6's and 8's you can never count on that to be true when you are playing, those numbers could just disappear when you are playing then come back when you lost all of your money!

I do not disagree with the math of the game, but I will never be there for 10,000 rolls in one session, and I will never just keep making a pass-line bet and nothing else. There aren't to many players that will ever play that way.

Any time you do a simulation you have to take into consideration how the player would actually play.
No players will be there for 10,000 rolls the dice, if they are smart they will only be at a craps table for an hour or two! So you have to base how you bet on what's happening on the table while you're playing.

RaleighCraps has a very good understanding of the game and bets accordingly to what's happening on the table when he's there! If you realize that the math of the game does not work out in the short period of time that you're at a table, you can adjust your play and your betting style to take advantage of what is happening.

If you're a firm believer in the math of the game, and that its infallible you should never place your first bet on a craps table! So my question always comes back to anybody that says no matter what you are doing, you are going to lose, just how often do you play craps in a real live casino?
Is it more fun for you to run your simulations or do you have fun continuously losing over what you consider the long run, never changing the way you are betting?

I've asked this question before of a few members on this board, did you win by making a pass-line bet and taking full odds when you where playing in Vegas when the WoV spring fling was in town, and the few guys that answered that question, all said no!

Having a predetermined betting strategy can be a death sentence to your bankroll if you never learned to adapt to the table conditions you are playing on.

superrick & RaleighCraps are saying that HA doesn't matter; it's whether or not your numbers hit, regardless of how you bet. I'm coming to that belief. If there is any skill in playing craps, it's the skill they describe: being aware of what numbers are showing, then bet accordingly, whether it's a "best bet" or a "sucker bet." But, it seems to me that that's a very difficult skill to develop, and it also predicts future performance, based on past performance. Commonly called "The Gambler's Fallacy."

I'm not denying what they say is true. From my brief experience at the tables, I've regretted more than once sticking to a PL full odds, and PL full odds & 1 Come full odds. I do believe a player should have more than 1 number in play or more, depending on the BR. Just yesterday, I decided to play PL full w/ 2x odds. (My house limits odds to 2x.) Other guys were placing & propping all over the felt & winning, so I decided to place either the 6 or 8 along w/ my PL bet. (My buy-in was & will be $200.)

After 3 sessions on the day, I was $5 ahead. All to the good, but along the way, it didn't matter how/what you bet, as long as your numbers were hitting. My 1st session was great. I walked after 1 hr., $145 to the good. (Should have stayed longer.) My last session (shouldn't have walked up to the table from the looks of it), I walked after 1/2 hr. because the table was bad. Even then, I lost $85.

SevenOut likes the "hit & run" strategy. There's also the "cut & run" strategy. But, it does get down to the skill of reading a table & betting or not betting, accordingly. Sometimes that skill falls short, but there you are.

I'm not saying HA doesn't matter, because it does. The more often you play on the worst bets, the more of an edge you are fighting against. This is important to understand.

What I am saying is don't let that be the ONLY factor that you consider when playing craps.
It is important that you know what you are up against with each bet that you make, so that you can make educated decisions.
You are also correct that past numbers have nothing to do with future numbers, at least not in any guaranteed way. However, I do feel that a table does trend, and if you catch a trend early, you can get lucky and win.


For me, I will not bet the horn, nor any craps. Nothing in the center of the table, but hardways, fire bet, or Small/Tall/All. I have thought about putting a bounty out there that says if anyone ever sees me make a horn bet, I will give them money, I feel that strongly against those bets. My friends all watch to try and call me out on this, but they have not been able to yet.